I think "stop writing code" is decent advice in principle, but is practically difficult. A few points: 1. If you have no domain experience or expertise, you sh

I think "stop writing code" is decent advice in principle, but is practically difficult. A few points: 1. If you have no domain experience or expertise, you shouldn't just stop writing code, you should stop what you're doing. A lot of startups fail before any code is written because the founders know little to nothing about the market or industry they're targeting. 2. While it is absolutely possible to sell software before it has been written, unless you're well-connected and have trusted relationships that will deliver customers on tap, to get meaningful conversations started, you will often need to have something to show, even if it's mockups or a basic prototype. 3. A lot of startups write too much code because they try to support use cases that they think are important. Ironically, some of these use cases are actually discovered through conversations with potential customers, which highlights the fact that it's not simply enough to talk to people you hope might become your customers; you need to know how to talk to them.